AP-Net: An atomic-pairwise neural network for smooth and transferable interaction potentials
| العنوان: | AP-Net: An atomic-pairwise neural network for smooth and transferable interaction potentials |
|---|---|
| المؤلفون: | Alexios Koutsoukas, C. David Sherrill, Steven A. Spronk, Daniel L. Cheney, Zachary L. Glick, Derek P. Metcalf |
| المصدر: | The Journal of Chemical Physics. 153:044112 |
| بيانات النشر: | AIP Publishing, 2020. |
| سنة النشر: | 2020 |
| مصطلحات موضوعية: | Physics, 010304 chemical physics, Artificial neural network, Computation, Intermolecular force, Ab initio, General Physics and Astronomy, Interaction energy, 010402 general chemistry, 01 natural sciences, Potential energy, 0104 chemical sciences, Test set, 0103 physical sciences, Partition (number theory), Pairwise comparison, Statistical physics, Physical and Theoretical Chemistry |
| الوصف: | Intermolecular interactions are critical to many chemical phenomena, but their accurate computation using ab initio methods is often limited by computational cost. The recent emergence of machine learning (ML) potentials may be a promising alternative. Useful ML models should not only estimate accurate interaction energies, but also predict smooth and asymptotically correct potential energy surfaces. However, existing ML models are not guaranteed to obey these constraints. Indeed, systemic deficiencies are apparent in the predictions of our previous hydrogen-bond model as well as the popular ANI-1X model, which we attribute to the use of an atomic energy partition. As a solution, we propose an alternative atomic-pairwise framework specifically for intermolecular ML potentials, and we introduce AP-Net—a neural network model for interaction energies. The AP-Net model is developed using this physically motivated atomic-pairwise paradigm and also exploits the interpretability of symmetry adapted perturbation theory (SAPT). We show that in contrast to other models, AP-Net produces smooth, physically meaningful intermolecular potentials exhibiting correct asymptotic behavior. Initially trained on only a limited number of mostly hydrogen-bonded dimers, AP-Net makes accurate predictions across the chemically diverse S66x8 dataset, demonstrating significant transferability. On a test set including experimental hydrogen-bonded dimers, AP-Net predicts total interaction energies with a mean absolute error of 0.37 kcal mol−1, reducing errors by a factor of 2-5 across SAPT components from previous neural network potentials. The pairwise interaction energies of the model are physically interpretable, and an investigation of predicted electrostatic energies suggests that the model ‘learns’ the physics of hydrogen-bonded interactions. |
| تدمد: | 1089-7690 0021-9606 |
| DOI: | 10.1063/5.0011521 |
| URL الوصول: | https://explore.openaire.eu/search/publication?articleId=doi_dedup___::256d4b48ff09799a93bfa65cba7ea3e9 https://doi.org/10.1063/5.0011521 |
| Rights: | OPEN |
| رقم الانضمام: | edsair.doi.dedup.....256d4b48ff09799a93bfa65cba7ea3e9 |
| قاعدة البيانات: | OpenAIRE |
| تدمد: | 10897690 00219606 |
|---|---|
| DOI: | 10.1063/5.0011521 |